¿ªÔÆÌåÓý

Date

Frequency calibration on a lab #bfo #calibration #frequency

 

Hi folks,

Assuming I have RF laboratory instruments available, what could be an ideal scenario to calibrate ubitx frequencies ?

Would it work to set a RF generator to 10,000,500Hz and expect to receive 500Hz when the ubitx is tuned at 10MHz?

Or would it be better to measure the clk2 frequency with a counter?

73

Rafael Pinto - PU1OWL


Re: Different Nano's

 

On Sun, Jan 3, 2021 at 06:39 PM, Shirley Dulcey KE1L wrote:
TL:DR version: don't buy a Nano Every to use in your Raduino?right now unless?you are prepared to do some software development.
Hi Shirley and all -?

I have got my own homebrew sketch working on the Nano Every fine business.? The one change I had to make was the interrupt handling for the rotary encoder.? Discussion and sample sketch are here:??[email protected] | Nano Every - Interrupt Handling - Rotary Encoder

This code will work reliably with either the Nano or the Nano every - it automatically detects which board it is compiling for and adjusts.?

I believe this will approach will work fine with the Raduino sketches - but I have not tried it.

The other useful benefit of the Nano Every is there is a second hardware serial port which I used for a Nextion connection.? It gets you away from the somewhat troublesome SoftwareSerial library limitations.

Dean
KK4DAS


Re: Different Nano's

 

Shirley does know these Arduino variants better than I, that last message was meant to be self-deprecating.
I sent that out before reading Shirley's post 85141, which hit the forum one minute prior to mine.

Jerry, KE7ER


On Sun, Jan 3, 2021 at 05:15 PM, Jerry Gaffke wrote:
I probably should defer to Shirley.
But instead I'll just say "yes, it would work", except that the pins are soldered to the wrong side of the board for plugging into the Raduino.
You want a board that does not come with the pins soldered in place.


Re: Problems with connecting to computer for digital

 

I should correct that - you can set DTR high in WSJTX but I haven't had as much luck with that.? I think it times out during the Arduino reset - but it still works fine after using HRD.? I am working another thread with developers in the Software Controlled Ham Radio group - perhaps those fine folks can help or some of the brilliant Arduino programmers on here s well.? I will say there is a lot of mythology around this because things seem to work for some folks - and then they don't.? I am very interested in getting this working perfectly for my own radio control sketch.

The one consistent behavior for me is as I described.? After running HRD (V5) - all the other programs connect just fine.

Dean


Re: Different Nano's

 

I probably should defer to Shirley.
But instead I'll just say "yes, it would work", except that the pins are soldered to the wrong side of the board for plugging into the Raduino.
You want a board that does not come with the pins soldered in place.

Otherwise, primary difference is this one has the more expensive FTDI USB-to-UART chip,
instead of the CH340 chip that the cheap clones do.
And the build quality is probably better.

Pretty much everybody in the forum is buying the $3 clones.
And that's what HFSignals ships with the uBitx.

Jerry, KE7ER


On Sun, Jan 3, 2021 at 04:35 PM, Will B wrote:
Would this one work?? According to their site, it's the 'classic' version: ? Again, I am so clueless about Arduino products but am learning through this group.


Re: Different Nano's

 

That's the authentic official Arduino Nano. It will certainly work. But clones from China are much cheaper, if you care.


On Sun, Jan 3, 2021 at 7:35 PM Will B <will.brokenbourgh2877@...> wrote:
Would this one work?? According to their site, it's the 'classic' version: ? Again, I am so clueless about Arduino products but am learning through this group.


Re: Different Nano's

 

Curious.
The reason I gave the "better" link is because the original link didn't work for me.
I refreshed to my "better" link and it seemed to work.? But now it doesn't.
I assume it has to do with tracking how links get traded.

Anyways, searching for the Mouser part number always works.
Which is why I posted it.

Jerry


Re: Different Nano's

Will B
 

Would this one work?? According to their site, it's the 'classic' version: ? Again, I am so clueless about Arduino products but am learning through this group.


Re: Ideas for "bulletproofing" the uBitx

 

¿ªÔÆÌåÓý

I¡¯m glad I asked about the Nano before I placed an order. ?Thanks to all for the input and comments.

Scot WB7AVU

?

From: [email protected] <[email protected]> On Behalf Of Jerry Gaffke via groups.io
Sent: Sunday, January 3, 2021 4:37 PM
To: [email protected]
Subject: Re: [BITX20] Ideas for "bulletproofing" the uBitx

?

Not having to insulate the RD's tab from ground is definitely a win.

> the RD generates less heat per watt than the IRF510

Are you saying the RD is significantly more efficient?
Do you have numbers?
Or perhaps this is your way of pointing out that heat transfer from the die is somewhat slower for a given delta-T.

> Less drive per watt out is good, 20 watts if you need it at 13.8V.

I think you will have to do a significant redesign of the uBitx to get 20W out at 13.8v,
starting with the transformer at T11.
Considerably more involved than just swapping in some RD's.

I can believe that less drive is needed, but I'm not sure how big a factor this is.
I doubt it is terribly significant.?
The uBitx final is limited in part by how much power is available from the driver stage,?
but those who have tried swapping in RD's in the past have not seen much improvement.
Has anybody done measurements of drive required for RD's vs IRF510's from 3.5 to 30mhz?

The primary advantage of an IRF510 is that it doesn't cost $25 (with shipping) when you blow a pair.
There are disadvantages, but they can be designed around.

Jerry, KE7ER


On Sun, Jan 3, 2021 at 02:47 PM, Slurm wrote:

Seems for continuous digital operation the RD might be preferred just for the fact that you can bolt it's tab directly to an aluminum radio's case,, giving a very large heat sink, add to that , the RD generates less heat per watt than the IRF510.? Those two things together seem important for digital operation and heat concerns in general.

Less drive per watt out is good, 20 watts if you need it at 13.8V. Nice that the Ubitx already has holes in the PCB for the RD. Being fairly new to the tech of radio I'd have to study the changes to the output transformer and ask do you have to change the output transformer if you keep the supply voltage to the finals at the 13.8V??

Comes down to what people need or want.? For any heat sink, get a big one, and get the best heat transfer paste you can find.? You can also run a fan, 24Volt fans run at 12v are quiet. I've never used a water cooled heat sink on a computer and ask those who know how sturdy and reliable they are, also how much power consumption and noise is acceptable.

Thanks for this discussion.

Slurm?


Re: Different Nano's

Will B
 

Jerry said: This might be a better web link:??
Mouser part number is??782-ABX00028
Thanks for that Jerry.? I usually remove the extra stuff from URLs before sharing/posting but Mouser seems to want it there.? The link as shown doesn't work for me.

Will - AF7EC


Re: Different Nano's

Will B
 

Shirley and Jerry, thanks so much for educating me.? I'm enjoying the discussions here and learning more every day! :-)

Will B - AF7EC


Re: Different Nano's

 

Ah, you are correct.
I did a quick read of the datasheet for the Nano Every and saw mention of an ARM processor, but it is not what our code runs on.
Seemed weird that they could get the Nano module pin compliant with such a totally different processor.

Jerry, KE7ER


On Sun, Jan 3, 2021 at 03:39 PM, Shirley Dulcey KE1L wrote:
TL:DR version: don't buy a Nano Every to use in your Raduino?right now unless?you are prepared to do some software development.
?
The ATMega 4809 is not an ARM processor. It's still an 8 bit AVR. It does have more RAM and flash memory than the ATMega328P does. The pinout of the 4809 is different from the 328P (it's not just a matter of having more pins), but the layout of the Nano Every board gets around that and puts the pins back where you expect them. Code changes may be needed; anything that uses libraries that come with the Arduino IDE is already covered, but the peripherals are different so any code that talks to ports directly or uses third party libraries is likely to need rewriting.
?
It's not quite a drop-in replacement, but adapting the Raduino?for the Nano Every should not be difficult in most applications. I don't know if anybody has done it yet for the?Raduino; until that happens, I don't recommend?that anybody make the switch unless they plan to do that work. The advantage would be the ability to use a larger sketch without the need to make the more extensive code changes needed to switch to another microcontroller such as the Teensy. There are two notable hardware differences that can affect compatibility?in some applications: PWM is not available on D11, and the USB serial interface is now completely separate from the serial interface on RX0 and TX1 so you can use both at the same time. (The USB to serial is Serial, the?other is Serial1.)
?
The Nano Every uses a different chip for USB to serial conversion: a pre-programmed?ATSAMD11D14A Cortex-M0 ARM processor. Ironically, that's more powerful in some ways than the board's primary CPU. The driver on Windows gets installed when you install the Arduino IDE; if you have an old version you may need to update it.
?
Here is a link to a blog post that talks about the hardware differences:?


Re: Different Nano's

 

The issue with resonators and such is that they tend to wander.
This can sometimes on a few radios put a birdie right in the middle of an IF passband.
The uBitx IF got moved from 12 to 11mhz because of a particularly nasty case of this.

Jerry


On Sun, Jan 3, 2021 at 03:12 PM, Shirley Dulcey KE1L wrote:
Reliable operation of an asynchronous serial link does not require a highly accurate clock; it only needs to be within 1% or so. That's easily met by a ceramic resonator. The CH340C is probably not even using one of those; it's using a reference on the silicon itself, perhaps helped out by a factory-programmed code for the specific chip. That's how the internal clock that is available on many microcontrollers works, including the ATMega328P which can be run without an external frequency reference.
Hide quoted text


Re: Different Nano's

 

TL:DR version: don't buy a Nano Every to use in your Raduino?right now unless?you are prepared to do some software development.

The ATMega 4809 is not an ARM processor. It's still an 8 bit AVR. It does have more RAM and flash memory than the ATMega328P does. The pinout of the 4809 is different from the 328P (it's not just a matter of having more pins), but the layout of the Nano Every board gets around that and puts the pins back where you expect them. Code changes may be needed; anything that uses libraries that come with the Arduino IDE is already covered, but the peripherals are different so any code that talks to ports directly or uses third party libraries is likely to need rewriting.

It's not quite a drop-in replacement, but adapting the Raduino?for the Nano Every should not be difficult in most applications. I don't know if anybody has done it yet for the?Raduino; until that happens, I don't recommend?that anybody make the switch unless they plan to do that work. The advantage would be the ability to use a larger sketch without the need to make the more extensive code changes needed to switch to another microcontroller such as the Teensy. There are two notable hardware differences that can affect compatibility?in some applications: PWM is not available on D11, and the USB serial interface is now completely separate from the serial interface on RX0 and TX1 so you can use both at the same time. (The USB to serial is Serial, the?other is Serial1.)

The Nano Every uses a different chip for USB to serial conversion: a pre-programmed?ATSAMD11D14A Cortex-M0 ARM processor. Ironically, that's more powerful in some ways than the board's primary CPU. The driver on Windows gets installed when you install the Arduino IDE; if you have an old version you may need to update it.

Here is a link to a blog post that talks about the hardware differences:?

On Sun, Jan 3, 2021 at 6:11 PM Jerry Gaffke via <jgaffke=[email protected]> wrote:
Hmm,? that's a genuine "Arduino Nano Every".
Has anyone tried that on a Raduino yet?
I think it could be made to work, but may require some changes to the source code.

This might be a better web link:??
Mouser part number is??782-ABX00028

This board is based on the ATMega4809, which is an Arm Cortex M0+ 32 bit processor, a 20mhz clock.
Whereas the uBitx comes with a clone of the original Arduino Nano, which has an ATMega328P,
which is a much less advanced 8 bit processor.
With 48k of Flash and 6k of RAM it has somewhat more than the 328P, though it's possible
that the Arm Cortex M0+ uses it up more quickly.
This genuine Arduino will likely have better quality control than the Nano Clones.
The pin out of this module appears to be exactly the same as the older Nanos.
Price at $11 is very reasonable.

So it is possible that it could plug directly into the Raduino, you recompile the Raduino source code
for the new processor under the Arduino IDE, and it all just works.
I am guessing there will be more work than that.
For example, the old code might run faster on the ARM, and this might expose a timing bug.
Or, even though it has SPI and I2C and UART and Power and GND and digital and analog pins
in all the right spots, there might be some minor difference about how an analog pin works
that hangs things up.? ?

Jerry, KE7ER


On Sun, Jan 3, 2021 at 02:22 PM, Will B wrote:
This is my first post to this group, and I'm more of a Raspberry Pi guy myself, so please be gentle. :-)

Would this work as a replacement?


Re: Ideas for "bulletproofing" the uBitx

 

Not having to insulate the RD's tab from ground is definitely a win.

> the RD generates less heat per watt than the IRF510

Are you saying the RD is significantly more efficient?
Do you have numbers?
Or perhaps this is your way of pointing out that heat transfer from the die is somewhat slower for a given delta-T.

> Less drive per watt out is good, 20 watts if you need it at 13.8V.

I think you will have to do a significant redesign of the uBitx to get 20W out at 13.8v,
starting with the transformer at T11.
Considerably more involved than just swapping in some RD's.

I can believe that less drive is needed, but I'm not sure how big a factor this is.
I doubt it is terribly significant.?
The uBitx final is limited in part by how much power is available from the driver stage,?
but those who have tried swapping in RD's in the past have not seen much improvement.
Has anybody done measurements of drive required for RD's vs IRF510's from 3.5 to 30mhz?

The primary advantage of an IRF510 is that it doesn't cost $25 (with shipping) when you blow a pair.
There are disadvantages, but they can be designed around.

Jerry, KE7ER


On Sun, Jan 3, 2021 at 02:47 PM, Slurm wrote:
Seems for continuous digital operation the RD might be preferred just for the fact that you can bolt it's tab directly to an aluminum radio's case,, giving a very large heat sink, add to that , the RD generates less heat per watt than the IRF510.? Those two things together seem important for digital operation and heat concerns in general.

Less drive per watt out is good, 20 watts if you need it at 13.8V. Nice that the Ubitx already has holes in the PCB for the RD. Being fairly new to the tech of radio I'd have to study the changes to the output transformer and ask do you have to change the output transformer if you keep the supply voltage to the finals at the 13.8V??

Comes down to what people need or want.? For any heat sink, get a big one, and get the best heat transfer paste you can find.? You can also run a fan, 24Volt fans run at 12v are quiet. I've never used a water cooled heat sink on a computer and ask those who know how sturdy and reliable they are, also how much power consumption and noise is acceptable.

Thanks for this discussion.

Slurm?


Re: Different Nano's

 

Reliable operation of an asynchronous serial link does not require a highly accurate clock; it only needs to be within 1% or so. That's easily met by a ceramic resonator. The CH340C is probably not even using one of those; it's using a reference on the silicon itself, perhaps helped out by a factory-programmed code for the specific chip. That's how the internal clock that is available on many microcontrollers works, including the ATMega328P which can be run without an external frequency reference.


On Sun, Jan 3, 2021 at 3:15 PM Jerry Gaffke via <jgaffke=[email protected]> wrote:
Banggood claims their Nano Clones to be "improved", though I doubt this matters to us.

Banggood gives this rather cryptic comment regarding the CH340 USB chip that is used:

####################
Chip: CH340C/G Random delivery?(CH340B is the update version of CH340G, CH340B comes with crystal, but CH340G doesn't
)
Note:??Improved USB chip for speed and stability using WIN7/WIN8.
####################

Here's something a bit more coherent from:? ??
Parts of that document look like a datasheet, others look like commentary about a bunch of different parts.
I have seen a CH340 datasheet in English, but perhaps some of the newer variants only
have datasheets in Chinese?? Anyways, here's the key passage from that document:

####################
Model differences: CH340C, CH340E and CH340B have built-in crystal, no external crystal; CH340B also has built-in EEPROM used to configure the serial number,etc.Some functions can be customized. CH340R provides reverse polarity TXD and MODEM signals.
####################

So perhaps banggood is now selling a mix of parts, some have the old CH340G chip with an external crystal,
some have a newer chip (either CH340B or CH340C) with the crystal inside the CH340 SOIC16 package?
Or perhaps they are now selling a mix of CH340B and CH340C, both of which have a crystal inside?

The CH340 chip is the 16 pin SOIC on the left side of the second photo in the banggood webpage.
A crystal can be seen just above and near the left end of the CH340 chip,
a long yellowish rectangle with traces to pins 7 and 8 of the CH340.
So the photo probably shows an older Nano Clone using the CH340G chip.

My understanding was that the Nano Clones are all using the cheap ceramic resonators
(12mhz on the CH340, 16mhz for the ATMega382P), not quartz crystals like the original Arduino Nano did.
I'm guessing what's inside a CH340B or CH340C is also a resonator of some sort,
so not a particularly stable or accurate source of 12mhz.

Jerry, KE7ER


On Sun, Jan 3, 2021 at 08:43 AM, Jerry Gaffke wrote:
Scot,

I doubt they have changed.
Perhaps you are looking at the bottom side of your board,
and the top side of the various product offerings?

Be sure to buy a Nano that does not already have the pins soldered,
because the raduino wants them installed upside down.

This webpage has images of top and bottom of the board:
??

Jerry, KE7ER


Re: Different Nano's

 

Hmm,? that's a genuine "Arduino Nano Every".
Has anyone tried that on a Raduino yet?
I think it could be made to work, but may require some changes to the source code.

This might be a better web link:??
Mouser part number is??782-ABX00028

This board is based on the ATMega4809, which is an Arm Cortex M0+ 32 bit processor, a 20mhz clock.
Whereas the uBitx comes with a clone of the original Arduino Nano, which has an ATMega328P,
which is a much less advanced 8 bit processor.
With 48k of Flash and 6k of RAM it has somewhat more than the 328P, though it's possible
that the Arm Cortex M0+ uses it up more quickly.
This genuine Arduino will likely have better quality control than the Nano Clones.
The pin out of this module appears to be exactly the same as the older Nanos.
Price at $11 is very reasonable.

So it is possible that it could plug directly into the Raduino, you recompile the Raduino source code
for the new processor under the Arduino IDE, and it all just works.
I am guessing there will be more work than that.
For example, the old code might run faster on the ARM, and this might expose a timing bug.
Or, even though it has SPI and I2C and UART and Power and GND and digital and analog pins
in all the right spots, there might be some minor difference about how an analog pin works
that hangs things up.? ?

Jerry, KE7ER


On Sun, Jan 3, 2021 at 02:22 PM, Will B wrote:
This is my first post to this group, and I'm more of a Raspberry Pi guy myself, so please be gentle. :-)

Would this work as a replacement?


Re: Different Nano's

 

There are the new Nano Every and Nano 33 from Arduino. Those are different, and are not 100% compatible with the earlier Nano. You don't want either of those unless you are prepared to make changes to the code.

Just about all of the clone Nanos you can buy now are not identical to the original Nano from Arduino; they use the CH340C or CH340G USB to serial chip instead of the FTDI FT232RL that is used on the official boards. Aside from markings on the chip, the difference is easy to spot because the CH340 is a smaller chip with fewer pins. The change is for cost reasons; you can buy an entire Nano clone for less than the price of that FTDI chip from a distributor. It's not an issue for most users; current versions of Windows automatically download the CH340 driver?the first time you plug one in, and Linux and macOS come with it built-in. (If you're still a Windows XP holdout you'll have to download and install the driver yourself, and I don't remember whether Windows 7 can automatically download the driver.)

The CH340G is a USB to serial chip from China. It costs 30 to 40 cents, while an FTDI chip is $4.50 in quantity one and $2.65 even if you buy 1,000. The CH340C is a newer variant that reduces costs even further; the chip costs about the same but has a built-in frequency reference, eliminating the need for a crystal. Many of the clone makers have now switched to the CH340C, but you may still find some with the CH340G. The ones with the CH340C won't have a crystal on the back of the board.

The clones look mostly the same on the top side (aside from trademark graphics) with one exception; most of them have an oval shaped reset button rather than the round button used on the official boards from arduino.cc

The BITX line has always?used non-official Nanos; the official ones are too expensive for the low cost goal of the project. But if you have one that's old enough it could have a slightly different design of board.

A Nano without the voltage regulator would not be a fully compatible replacement. It would only be suitable in applications where the board is powered by the USB port. I have never seen one for sale. However, I did find one blog that shows a Nano clone with the regulator removed; the owner used a boost regulator and a LiPo battery to power it.

Here are pictures of the back of a standard Nano, and of the CH340G variant with the regulator removed. The layout of the back is substantially different on the CH340G version.Although the chip markings on the second picture are barely visible, you can tell it's a CH340G rather than a CH340C because the board has a crystal.

Arduino Nano v3_2-1000x750.jpg
arduino-nano-1117-removed.jpg

On Sun, Jan 3, 2021 at 11:33 AM Scot McMath <scotmcmath@...> wrote:
Looking to get some more Nano's, however when shopping on Amazon, E-Bay and others, They look different. Don't see the 5 volt regulator for 1 among offer things. Have the Nano's changed? and if so, can tey be used in my current V6?
Thanks for any info, Scot WB7AVU


Re: Different Nano's

 

That's a Nano Every. Not suitable as a direct replacement. It uses a different microcontroller and requires some code changes.


On Sun, Jan 3, 2021 at 5:23 PM Will B <will.brokenbourgh2877@...> wrote:
Greetings!

This is my first post to this group, and I'm more of a Raspberry Pi guy myself, so please be gentle. :-)

Would this work as a replacement?

Will B - AF7EC


Re: Ideas for "bulletproofing" the uBitx

Slurm
 

Seems for continuous digital operation the RD might be preferred just for the fact that you can bolt it's tab directly to an aluminum radio's case,, giving a very large heat sink, add to that , the RD generates less heat per watt than the IRF510.? Those two things together seem important for digital operation and heat concerns in general.

Less drive per watt out is good, 20 watts if you need it at 13.8V. Nice that the Ubitx already has holes in the PCB for the RD. Being fairly new to the tech of radio I'd have to study the changes to the output transformer and ask do you have to change the output transformer if you keep the supply voltage to the finals at the 13.8V??

Comes down to what people need or want.? For any heat sink, get a big one, and get the best heat transfer paste you can find.? You can also run a fan, 24Volt fans run at 12v are quiet. I've never used a water cooled heat sink on a computer and ask those who know how sturdy and reliable they are, also how much power consumption and noise is acceptable.

Thanks for this discussion.

Slurm?